GPU-enhanced Neuroscience Software Tools

描述（由申请人提供）：该应用程序旨在为神经科学家提供高性能、支持GPU的计算和可视化软件工具。今天，估计有150万生命科学MATLAB用户，其中很大一部分使用MATLAB解决神经科学相关问题。MATLAB用户，特别是那些处理大型神经科学数据集的用户，如脑MRI，fMRI，DW-MRI，PET和CT图像体积，显微镜图像和基因组学数据集，目前在使用MATLAB进行神经科学研究时存在两个主要问题：1）MATLAB与其他编程语言（如C/C++）相比速度较慢，以及2）MATLAB可视化不能处理大量数据或容易地呈现解剖结构的3D模型。因此，神经科学家经常花费大量的时间和精力将神经科学MATLAB代码移植到C/C++中，代价是减缓研究工作，合作，并最终偏离研究人员解决生物学问题的主要重点。基于计算机处理器的最新进展，特别是由于NVIDIA的Tesla，AMD的Firestream和英特尔即将推出的多集成核心（MIC）处理器，新一波的处理技术使个人研究人员能够直接在MATLAB中获得更快的速度和更强的可视化。在过去的四年中，我们开发并发布了我们的第一款产品Jacket：MATLAB的GPU引擎，它使科学家能够在GPU上执行低级MATLAB计算。在第一阶段，我们成功地在GPU上加速了神经科学家常用的一组构建块MATLAB函数，例如MATLAB的信号处理，图像处理和统计工具箱中的函数。在第二阶段，我们计划利用第一阶段的成功，为MATLAB社区提供更全面的GPU增强神经科学功能套件。通过对Jacket用户社区的各种调查，我们已经确定了在MATLAB神经科学社区中取得研究进步所需的3项主要能力：更快的医学图像处理，更快的生物信息学算法以及利用状态的可视化功能。 直接在MATLAB中绘制图形。 公共卫生相关性：该项目的目的是推进Jacket的开发，为神经科学家提供高性能的GPU支持的工具。今天，估计有150万生命科学MATLAB用户，其中很大一部分使用MATLAB解决神经科学相关问题。MATLAB用户，特别是那些处理大型神经科学数据集的用户，如脑MRI，fMRI，DW-MRI，PET和CT图像体积，显微镜图像和基因组学数据集，目前在使用MATLAB进行神经科学研究时存在两个主要问题：1）MATLAB与其他编程语言（如C/C++）相比速度较慢，以及2）MATLAB可视化不能处理大量数据或容易地呈现解剖结构的3D模型。由于计算机处理器的最新进展，特别是由于NVIDIA的Tesla，AMD的Firestream和英特尔即将推出的Many Integrated Core（MIC），新一波的桌面和服务器处理器技术使得个人研究人员可以直接在MATLAB中获得更快的速度和更强的可视化。在这项工作中，我们将通过GPU扩展Jacket-启用流行的统计参数映射和生物信息学，并通过增强我们的医学成像和生物信息学可视化库。

DESCRIPTION (provided by applicant): This application is to deliver high-performance, GPU-enabled computation and visualization software tools to neuroscientists. Today, there are an estimated 1.5 million life science MATLAB users, with a substantial portion of those using MATLAB to solve neuroscience-related problems. MATLAB users, especially those dealing with large neuroscience datasets, such as brain MRI, fMRI, DW-MRI, PET, and CT image volumes, microscopy imagery, and genomics datasets, currently have two major problems in using MATLAB to conduct neuroscience research: 1) MATLAB is slow when compared to other programming languages such as C/C++, and 2) MATLAB visualizations are unable to handle large amounts of data or to render 3D models of anatomical structures with ease. Therefore, neuroscientists often undertake costly and time-consuming efforts to port neuroscience MATLAB code to C/C++, at the expense of slowing down research efforts, collaborations, and ultimately detracting from the researcher's primary focus of solving biological problems. Building upon recent advances in computer processors, specifically due to NVIDIA's Tesla, AMD's Firestream, and Intel's upcoming Many Integrated Core (MIC) processors, a new wave of processing technology makes it possible for individual researchers to get increased speed and enhanced visualizations directly in MATLAB. Over the last four years, we have developed and released our first product, Jacket: The GPU Engine for MATLAB, which enables scientists to perform low-level MATLAB computations on the GPU. In Phase I, we were successful at GPU accelerating a set of building block MATLAB functions commonly used by neuroscientists, such as those found in MATLAB's Signal Processing, Image Processing, and Statistics Toolboxes. In Phase II, we plan to leverage the success of Phase I to deliver a more comprehensive suite of GPU-enhanced neuroscience functions to the MATLAB community. Through various surveys of the Jacket user community, we have identified 3 primary competencies that are needed to make research advancements in the MATLAB neuroscience community: faster medical image processing, faster bioinformatics algorithms, and visualization capabilities that leverage state-of the-art graphics directly in MATLAB. PUBLIC HEALTH RELEVANCE: The purpose of this project is to advance the development of Jacket to deliver high performance GPU- enabled tools to neuroscientists. Today, there are an estimated 1.5 million life science MATLAB users, with a substantial portion of those using MATLAB to solve neuroscience-related problems. MATLAB users, especially those dealing with large neuroscience datasets, such as brain MRI, fMRI, DW-MRI, PET, and CT image volumes, microscopy imagery, and genomics datasets, currently have two major problems in using MATLAB to conduct neuroscience research: 1) MATLAB is slow when compared to other programming languages such as C/C++, and 2) MATLAB visualizations are unable to handle large amounts of data or to render 3D models of anatomical structures with ease. Due to recent advances in computer processors, specifically due to NVIDIA's Tesla, AMD's Firestream, and Intel's upcoming Many Integrated Core (MIC), a new wave of desk-side and server processor technology makes it possible for individual researchers to get increased speed and enhanced visualizations directly in MATLAB. In this work, we will extend Jacket by GPU- enabling the popular Statistical Parametric Mapping Toolbox and the Bioinformatics Toolbox and by enhancing our visualization library for medical imaging and bioinformatics.